4.2 Export Particle Data to Blueprint【輸出粒子數(shù)據(jù)到藍圖】

In order to tell our Niagara System which blueprint to use, we first make a User Parameter in the? "System Settings" section of Niagara.

This User Parameter will be of the type "Object" and can be named whatever you like. In this case, the name is BPCallBackHandler.

【為了告訴Niagara系統(tǒng)哪一個藍圖使用數(shù)據(jù)，我們首先在Niagara系統(tǒng)中設置一個用戶參數(shù)】

【此用戶參數(shù)為Object【對象】類型，可隨意命名。在此處它的名字為BPCallBackHandler】

Click Class Settings, then add the Niagara Particle Callback Handler under "Interfaces"

【點擊Class Settings（類設置）（在藍圖界面的工具欄），在細節(jié)面板的Interface（接口）分類點擊Add添加NiagaraParticleCallbackHandler（Niagara粒子回調(diào)處理器）】

Then, In your "My Blueprint" tab, right click the "Receive Particle Data" Interface, and choose "Implement". This callback hander will appear.

【然后，在My Blueprint（我的藍圖）窗口，右鍵Receive Particle Data（接收藍圖數(shù)據(jù)）接口，選擇Implement Event（執(zhí)行事件），此回調(diào)處理器就會顯示（就會出現(xiàn)Event Receive Particle Data（接收粒子數(shù)據(jù)事件）藍圖節(jié)點）】

This event runs once every tick, but only if there was data sent from at least one particle.

【此事件每幀運行一次，但只有當有數(shù)據(jù)從粒子處發(fā)送時才會執(zhí)行事件】

Each Particle that sends data across packs their information into an array of structs, and can be iterated over.

【發(fā)送數(shù)據(jù)的粒子把它們的數(shù)據(jù)打包成一個結構數(shù)組，可被迭代】

The struct has named elements, but you can send whatever data you want across. You get two vectors and a float.

【此結構有已命名的元素，但你可以發(fā)送任何你想要的數(shù)據(jù)。你獲得兩向量和一浮點】

Do whatever blueprint work you like. Sky's the limit!

Note, if you are trying to play sound per particle, we also have dedicated Play Audio modules which streamline that behavior.

【可以使用該數(shù)值執(zhí)行任何藍圖行為，無任何限制】

【注：如果你想每個粒子播放聲音，我們有專門的播放聲音模組可簡化此行為】

GPU Particle Data Export is performed via a readback from the GPU memory. This is unpredictably latent by a non-fixed number of frames, usually 1-2.

【GPU粒子數(shù)據(jù)輸出是通過GPU內(nèi)存的回讀執(zhí)行的。這通常在1或2幀后執(zhí)行，這是不可確定的】

GPU readback performance is constrained by the size of the readback buffer. GPU fixed size allocation is recommended, with the smallest value necessary to capture all the events that could happen on a given frame.

In this example we know based on spawnrate and lifetime that we will only ever have ~4 particles alive at any given time, so the fixed size can be quite small.

Any events which exceed this fixed size will simply fail to send.

【GPU回讀性能被回讀緩存的大小所限制。建議使用GPU固定大小分配量，為在任一幀中能捕獲所有事件所必需的最小值?！?/p>

【在此示例，我們基于生成速率和壽命可知任一時間最多存在4個粒子，所以這固定大小可以非常小】

【任何超過固定大小的事件將簡單地發(fā)送失敗】

Export Particle Data to Blueprint【輸出粒子數(shù)據(jù)到藍圖】模組

Condition To Export Data【輸出數(shù)據(jù)的條件】：為True時才輸出數(shù)據(jù)

Delay Between Data Exports【輸出數(shù)據(jù)間延遲】：設置兩次輸出數(shù)據(jù)的最小間隔時間

Vector to Sent（As Struct Position）【發(fā)送的向量（被命名為位置）】

Vector to Send（As Struct Velocity）【發(fā)送的向量（被命名為速度）】

Float to Send（As Struct Size）【發(fā)送的浮點數(shù)（被命名為大?。?/p>

Callback Handler Parameter【回調(diào)處理器參數(shù)】：選擇把藍圖傳進Niagara系統(tǒng)的Object類型用戶參數(shù)

GPUAllocation Mode【GPU分配模式】：僅GPU粒子需要設置，建議使用Fixed Size【固定大小】

GPUAllocation Fixed Size【GPU分配固定大小】：一幀中發(fā)送數(shù)據(jù)的粒子數(shù)量的最大值，超過此數(shù)值的粒子無法發(fā)送事件。